TWI586944B - High throughput hot testing method and system for high-brightness light-emitting diodes - Google Patents

Description

High-volume heat test method and system for high-brightness light-emitting diode

The present invention relates to a method for rapidly establishing an operational solid-state LED product representing a high-intensity light-emitting diode (HBLED) or a phosphor-converted HBLED (pc-HBLED) (hereinafter referred to as HBLED) including a package. A method of conditional thermal test conditions. The present invention is also directed to a system that provides such high thermal measurement conditions and high precision measurements of the optical properties of such HBLEDs.

This application claims the application of the US Provisional Patent Application No. 61/559,411, entitled "HBLED High Throughput Hot Testing Method And Instrument", and the application titled "HBLED High Throughput" on November 16, 2011. The priority of U.S. Provisional Patent Application Serial No. 61/560,614, which is incorporated herein by reference.

1 illustrates an exemplary HBLED 100 comprising a phosphor layer 102 and a thin (e.g., a plurality of thick) indium gallium nitride (InGaN) film 101. In a typical embodiment, the phosphor layer 102 is also applied to the side of the InGaN film 101. Phosphor layer 102 comprises a luminescent phosphor (i.e., a microcrystal containing active visible luminescent ions). Phosphor layer 102 further comprises a binder (such as polyfluorene oxide) or a sintered crystal (described in more detail below). The InGaN film/phosphor layer combination is mounted on a submount 104 and then encapsulated using a lens 103 having a radius of about 2 mm. The lens 103 is formed using a polyfluorinated oxygen that increases light extraction by closely matching the refractive index of the surface of the phosphor layer 102.

The HBLED 100 can be inspected at the wafer level that is not singulated or singulated at different stages of processing. When the HBLED 100 is assembled into a product grade HBLED, the submount 104 can be further attached to a printed circuit board and a heat sink.

In general, the photometric parameters of the HBLED are measured during the electrical probe test. Exemplary photometric parameters include CCT (correlated color temperature, that is, the appearance of the emitted light and the combination of red, orange, yellow, white, and blue light to different degrees to form white in various positions along the Planckian curve One of the metrics relating to the appearance of a black body heated by one theory, chromaticity (regardless of the quality of one of its illuminances, ie, as determined by its hue and chroma: saturation, chroma, intensity or excitation purity) and CRI (color rendering index, i.e., using the average color of eight standard test score R _i R96 or the like, such as _a color test and the relevant test of the CIE (international Commission on illumination) system of the primary measurement). Probe testing is typically performed by applying a short current pulse to the InGaN film while measuring the optical properties of the HBLED for a time frame typically between 10 milliseconds and 200 milliseconds. Alternatively, the electrical probe can be applied for a period of time beyond the parameter measurement in an attempt to bring the thermal conditions of the HBLED to conditions that are more closely related to the conditions expected in the final lighting product form.

Unfortunately, the use of electrical probes does not bring the HBLEDs close to any of the conditions expected in the final lighting product. The main difficulty is attributed to the dissimilar thermal qualities of the materials used in the construction of HBLEDs. Such materials may comprise an InGaN film, a tantalum or copper submount, a sapphire or SiC substrate, a quartz material for a lens or an optical window, and a polymer for applying a slurry of a microcrystalline paste. A cerium encapsulating agent, wherein the microcrystals contain a luminescent material such as cerium and trivalent cerium ions. Of particular note is an organopolyphosphorus phosphor carrier paste having a thermal conductivity that is at least two orders of magnitude lower than the other listed materials and, in most cases, three orders of magnitude lower, This produces a solid thermal time constant or response time that is approximately three orders of magnitude longer than other materials.

For example, the InGaN film 101 has an operating temperature (85 ° C) of about 60 ° C above room temperature, while the phosphor layer 102 can reach approximately 200 ° C when containing polyfluorene or 200 in certain regions of certain products. One operating temperature above °C. Note that the thermal time constant (i.e., the time to reach thermal equilibrium) is approximately 10 milliseconds for the InGaN film 101 and may be from one second to two seconds or more for the phosphor layer 102. For different regions of the HBLED 100, not only the thermal time constants are different, but also the heat capacities of the various regions vary depending on the size and volume. The heating phosphor layer 102 is slower than the InGaN film 101 for two reasons.

The measurement of the optical parameters by means of the materials at a temperature substantially different from the expected operating temperature of the product of the final package of a full luminaire produces incorrect results due to the emission wavelength and efficiency of the InGaN film 101. (Strength) is moderately temperature dependent. More importantly, the absorption and emission spectra of the active phosphor ions in the phosphor layer 102 and the quantum yield of stokes shifted emission radiation are also temperature dependent. Therefore, it is important to measure and report the photoluminescence properties of HBLEDs (their properties related to the response of the human eye) as close as possible to the conditions expected in the final product.

Note that using a current from a probe to generate an InGaN emission will bring the InGaN film 101 to operating temperature in about 10 milliseconds. The emitted blue radiation from InGaN is absorbed by the active ions in the phosphor of the phosphor layer 102, which are due to the Stokes shift of the phosphor absorption and emission wavelengths and the temperature dependent each phosphor. Non-radiative decay produces red or green or yellow (usually (Eu+2) or (Ce+3) radiation from various matrix materials) and waste heat in the microcrystals of the phosphor binder. However, the slow thermal response of the surrounding polyfluorene in the phosphor layer 102 requires that the active phosphor ions can completely bring the surrounding polyfluorene to nearly 200 in the surface adjacent to the lens 103 (ie, furthest from the InGaN film). °C is expected to maintain a full second to two second excitation before balancing the operating temperature.

However, applying electrical excitation to the InGaN film 101 for one second or two seconds is commercially unattractive. Specifically, a high throughput tool needs to complete its measurement of the InGaN film in approximately 50 milliseconds to process approximately seven 4 inch wafers per hour (each containing approximately 10,000 grains). Additionally, applying a current for this duration heats the InGaN film 101 well above its expected product operating temperature because the thermal capacity of the film substrate is insufficient to absorb the applied energy. To address this heating issue, the HBLED 100 can be attached to an aluminum heat sink by, for example, adding one of a thermal capacity and convectively cooling one of the overall structures of the product grade HBLED. However, this means that each HBLED 100 is singulated prior to its testing and then substantially packaged close to the product form. Thus, instead, only the film submount at the wafer level is attached to a film frame carrier and thus can only be exposed to current for 10 milliseconds to 20 milliseconds before exceeding its desired operating temperature. Therefore, the heating energy is applied by the electric current. The HBLED 100 is not sufficient to produce the desired operating conditions in the product grade HBLED. Therefore, there is no accurate, commercially viable thermal test of HBLEDs on the market today.

To address this shortcoming, the LED industry has used alternative testing. For example, in one test mode, an InGaN film is fabricated at the wafer level and then mounted on a common 1.6 mm thick alumina submount at the wafer level. Next, the alumina sub-substrate having a wider array (array mounted) is covered with a phosphor film distributed in a polyoxymethylene resin binder. The mounted arrays are then placed in a furnace to a temperature of about 85 °C. Then power the installed array for about 10 milliseconds. Note that within this 10 milliseconds, the InGaN film and phosphor regions remain at approximately 85 °C. At this time, the emission spectrum and the average CIE coordinate of the entire alumina sub-substrate were recorded under the condition that all regions were nominally 85 °C.

Unfortunately, color coordinate measurement accuracy is sacrificed in the case of such temperature conditions for measurement. In particular, it is well known that the temperature of the phosphor layer varies from about 85 ° C in the vicinity of the InGaN film to a temperature of up to about 200 ° C in the region in contact with the thickest portion of the lens in the final product. The higher temperature in the hottest region of the phosphor layer causes an increase in the non-radiative decay of the active ions and a decrease in the conversion of the blue pumped radiation to the longer wavelength phosphorescence, thus significantly shifting the final color coordinates of the HBLED (see Figure 2). ). The exact amount of this shift is affected by the thickness of the phosphor, the level of phosphor doping, the type of phosphor, and additional factors. Many MacAdam elliptical shifts occur when the HBLED operating conditions are changed from room temperature to 85 °C. The displacement accelerates at a higher temperature.

As is known to those skilled in the art, a MacAdam ellipse is an elliptical region centered on one of the target colors on a chromaticity diagram. Each ellipse defines a color difference that becomes a threshold that the human eye perceives. The size of the MacAdam ellipse is stepped, where any point on the boundary of a 1-step MacAdam ellipse represents a standard deviation of the human perception of color mismatch between the two test samples. Matching the color of the approximately 2 step MacAdam ellipse is generally considered desirable for high quality lighting applications. The chromatic aberration of a larger number of MacAdam ellipse is not expected as a high quality illumination in achromatic side-by-side lighting applications. Therefore, the size of the product grading box of five MacAdam ovals (and some sticking to as few as three ovals) is not commercially attractive.

A relatively accurate technique for grading and color control involves the use of a phosphor plate that has been carefully selected and combined with different illuminating emitters to achieve a two-step MacAdam elliptical grading bin. An adjustable screw of the phosphor converter can also be inserted into each device and tuned by moving its position at the top of the light generating chamber to manually align the bulb color coordinates. These hand-made LEDs are constructed one at a time in a relatively high cost manufacturing process involving one of the trial and error methods. At present, it is difficult to achieve a product classification box of 3 to 4 MacAdam Ellipse in size even using this intensive manufacturing technique.

As demonstrated above, current manufacturing processes are not capable of achieving fast, accurate, commercially viable testing and grading for lighting applications. Accordingly, there is a need for an improved method to produce operating conditions similar to those of a product grade HBLED, thereby allowing for customer related measurements.

The invention describes a method for performing a high brightness light emitting diode (HBLED) The method of thermal testing. The HBLED includes an InGaN film, a phosphor layer formed on the InGaN film, and a lens formed on the phosphor layer and above the InGaN film. The method includes selectively using a laser to heat a portion of the phosphor layer to provide a predetermined temperature gradient in the phosphor layer. A current is applied to the InGaN film to provide a predetermined temperature within the InGaN film. After the selective heating and during the applied current, once the InGaN film temperature is established, the photometric measurement is performed on the HBLED.

Selective heating can be directly heat a phosphor based layer poly silicon-oxygen of the poly-silicon oxide or directly heating a phosphor based layer Lumiramic ^TM's in the active phosphor (containing sodium of the active within a sintered ceramic matrix material, the ionic composition of one Phosphor). Similarly, the laser can directly heat the active ion phosphor in the phosphor bonded phosphorous. In one embodiment, selective heating can be performed by means of a mid-infrared (middle IR) laser to directly heat the polyoxynoxy binder. In another embodiment, selective heating of the reactive ion phosphor can be performed in a polyoxonium or Lumiramic based phosphor layer by means of a visible tunable laser. In still another embodiment, selective heating may be performed by means of an optical pumped semiconductor laser (OPSL) or an InGaN laser diode array to excite at 450 nm or at a nominal 465 nm to 485 nm region. Absorption band. Tunable lasers, such as dye lasers, can also be used in the 450 nm or 465 nm to 485 nm region.

A system for thermal testing of HBLEDs in wafer level packaging includes a laser, a probe tester, an integrating sphere, and a spectrometer system. A laser positioned to direct its light onto an HBLED is configured to selectively heat a portion of the phosphor layer (eg, heat the adhesive (IR) or the phosphor itself ( see)). In one embodiment, the laser is positioned to direct its light through the integrating sphere onto the HBLED. The probe tester is configured to apply a current to the InGaN film of the HBLED to obtain a predetermined temperature within the InGaN film and to excite the phosphor during color coordinate measurement. The integrating sphere is configured to collect the light emitted by the HBLED during the test. The spectrometer system is configured to perform photometric measurements on the light collected by the integrating sphere. The system can further include timing electronics coupled to the laser and probe tester to synchronize the operation of the laser and probe tester. The system can further include a movable wafer carrier for positioning the HBLED.

The integrating sphere may include an optical ring positioned directly above the HBLED that is configured to maximize the collection of light from the HBLED to the integrating sphere at high angles (eg, from 10 degrees to 170 degrees) during testing. . This optically reflective ring can be used to ensure that all HBLED light is collected during color coordinate measurement.

The present invention illustrates another method of performing a thermal test of a HBLED. The method includes using a first excitation source to establish one of a first predetermined operating condition of the phosphor binder and a second excitation source to establish one of the second predetermined operating conditions of the die. Light metrics for the HBLEDs can be performed after using the first and second excitation sources. Establishing the first predetermined operating condition can include providing a predetermined temperature gradient for the phosphor binder, and establishing the second predetermined operating condition can include providing a predetermined temperature for the InGaN film.

Using the first excitation source can include targeting the excitation of the polyphosphonium or active phosphor ions in the phosphor layer. For example, in one embodiment, using the first excitation source can include using an optical light source to selectively excite the vibration mode of the polyoxygen oxide in the phosphor layer, thereby producing in the phosphor binder A temperature gradient is generated. In another embodiment, using the first excitation source can include using an optical source to selectively excite the vibration mode of the methanol or a hydrocarbon wetting agent in the phosphor layer, thereby creating a temperature gradient in the phosphor layer. The second excitation source can include applying a current to the InGaN film.

In one embodiment, one of the first excitation sources may have a wavelength between 2.0 microns and 3.5 microns, and one of the homologous sources may have an average power between 100 watts and 12 watts for selectively exciting the phosphor Any of a polysulfide doping wetting agent and a methanol doping wetting agent for the binder. In another embodiment, the first excitation source has a wavelength between 0.45 micrometers and 0.53 micrometers, and one of the homologous sources has an average power between 100 watts and 12 watts. The system can further include a plurality of first excitation sources that provide one or more of the average coherent power of 12 watts or more in combination.

The second excitation source can include an electrical probe tester. In an embodiment, the system can further include a plurality of second excitation sources.

The thermal diffusion time in a polyphosphonium-based phosphor layer and a sintered ceramic-based phosphor layer has a characteristic diffusion time of about several milliseconds and 100 microseconds, respectively, for a path length of about a fraction of the thickness of a phosphor. . Therefore, the temperature distribution established in the phosphor layer maintains only its initial amount curve or condition for a period of time before being degraded by diffusion into both the InGaN film and the lens close to the phosphor layer. Therefore, a gating spectrometer is used for each phosphor layer type to obtain color coordinates and luminance measurements during these characteristic times.

According to one aspect of a modified HBLED test method, a laser is used to preheat the phosphor or binder prior to applying a current to the InGaN film and then obtaining a photometric measurement of the wafer level packaged HBLED. When polyoxymethylene is used as a binder in the phosphor layer, the laser directly preheats the polyoxo, which rapidly allows the phosphor to reach the desired temperature profile in the final product. In contrast, when a phosphor Lumiramic ^TM (Proprietary one sintered ceramic phosphor plate is introduced by the Philips Lumileds company) was used as a phosphor layer, a direct heating laser active ions in the phosphor, so that the active ions subsequently fast Lumiramic ^TM based on the phosphor layer reaches the intended final product in the temperature distribution. Alternatively, in the case of a polyoxo-based phosphor layer, laser excitation can be used to directly heat the active ions in the phosphor to establish the desired temperature profile in the final product. In any of these embodiments, the subsequent optical metrology is an accurate representation of the performance within the product-level HBLED. Photometric measurements include CIE coordinates, CCT (correlated color temperature), chromaticity, and CRI.

Several factors contribute to the change in CIE coordinates within a given product batch. These factors include phosphor particle size distribution, morphology, and particle shape according to the Mie scattering theory. The non-uniformity of the phosphor mixed concentration, the hot spot existing on the InGaN film, and the thickness of the InGaN film and the overall phosphor layer (and thus the phosphor temperature distribution and electroluminescence conversion efficiency from blue to green and red wavelengths) Changes in forward voltage and efficiency (and therefore temperature) also contribute to CIE coordinate changes. The magnitude of the effect of some of these factors is detailed below.

The effectiveness of the InGaN film and its electroluminescence and the degree of non-radiative quenching of phosphorescence vary strongly with temperature and directly affect the ratio of blue/red/green light in an LED. This performance directly affects the CIE coordinates. Therefore, it must The precise operating temperature at each point in the phosphor is obtained to properly measure the product CIE coordinates and CCT.

Those skilled in the art have quantified changes in the efficacy of samples of phosphors that vary with temperature, and it has been noted that significant phosphorescence emissions change based on temperature changes. For example, Figures 2A and 2B use two different transmit filters to illustrate four phosphors that vary with temperature (specifically, (Y _1-x Ce _x ) ₃ (Al _1-y Ga _y ) ₅ O ₁₂ phosphor) performance. Figure 2A shows the change in 540 nm radiation decay time versus temperature, and Figure 2B shows the change in 700 nm radiation decay versus temperature in these phosphors. First, the photon conversion efficiency from blue to 540 nm or 700 nm radiation is proportional to the ratio of the decay time at elevated temperatures to the decay time at lower (typically 25 °C) temperatures. As described in Figures 2A and 2B, the phosphorescence emission (decay time) is varied by up to one factor or two factors as the temperature varies between 0 °C and 120 °C. Some phosphors are even more temperature sensitive, with even greater temperature dependence being measured between room temperature and temperatures above 400 °K and 400 °K. These temperature changes can produce substantial shifts in the CIE coordinates.

In addition, changes in a given phosphor and phosphor mixture within the same batch are affected by impurities, phosphor layer thickness, phosphor doping non-uniformity, and irregular phosphor distribution. For example, Figure 3 illustrates the variation in CIE coordinates as a function of phosphor doping levels for different sample phosphors. Note that GR11 has an equal portion of green phosphor and red phosphor; GR21 has two portions of green phosphor and one portion of red phosphor; GR31 has three portions of green phosphor and one portion of red phosphor; GR41 has 4 parts of green phosphor and one part of red Color phosphor. As shown by Figure 3, the weight ratio of green phosphor to red phosphor is changed from 4 to 1 (80%) (GR41) to 3 to 1 (75%) (GR31) to vary y by 0.08 and to vary x. A CIE shift of up to 0.025. 0.005 or better on x and/or y (which would be equivalent to a region of a MacAdam ellipse) would be optimal in the industry. However, as indicated by Figure 3, a change in weight percentage greater than 1.2% is sufficient to cause an unacceptable CIE shift.

It is also known that the CIE change is also attributed to variations in the thickness of the phosphor layer and the "bulk density", that is, the fraction of the phosphor binder comprising the phosphor as opposed to the binder. For example, Figure 4 shows an example of a CIE change in an exemplary dispensed phosphor layer.

The CIE coordinates of the yellow, red, and orange phosphors have also been measured as the drive current and/or surface mount temperature changes. Figure 5 shows CIE x vs. CIE y (shown as black to white circles) of a phosphor-powder mixture that is red-orange-yellow with a current from 800 mA to 100 mA as a function of temperature from 25 °C to 85 °C. A chart. The ellipse shown in Figure 5 represents 5 SDCM (standard deviation color matching) of the color from the center of the ellipse, in which case the center x is 0.42. As shown in Figure 5, the change in CIE coordinates when changing the surface mount temperature (the phosphor also undergoes heating from the Stokes shift of the absorbed light) corresponds to two MacAdam ellipse. Note that an additional one of the MacAdam elliptic variations between the grains at the same surface mount temperature can also occur. Therefore, the three ellipses of the color point change can be caused by the measurement under the erroneous temperature and the inconsistent production uniformity, and the color point change is significantly changed.

All in all, there are many reasons for the change in CIE. Therefore, for a production Inter-granular thermal testing is important in measuring and identifying CIE coordinates in a MacAdam ellipse under the final operating conditions within the product. Due to the multivariate nature of grain phosphor performance, it is common knowledge in the industry that measurements at one temperature and extrapolated to another do not produce accurate results. Therefore, thermal testing must be performed under the correct final product operating conditions.

According to an aspect of the improved HBLED thermal testing technique, the wafer has been processed to include an InGaN film, a phosphor layer, a lens, and a sub-substrate (see, for example, Figure 1) (generally, referred to in the industry as "bricks" After the block, the array of HBLEDs (or bricks) can be tested for light metrology. There are two main factors that cause difficulties in thermal testing. One of the first factors mentioned above is a substantial difference in the properties of the material used in the HBLED (specifically, its thermal properties). For example, a pn junction is typically formed using an InGaN film, and as an example, the substrate material is typically formed using sapphire or Al ₂ O ₃ or tantalum carbide. In general, Ce ⁺³ (铈) or Eu ⁺² (铕) or related active ions are embedded in a plurality of microparticle crystalline matrices (such as YAG (yttrium aluminum garnet), CaS (calcium sulfide), Ca ₁ embedded in the phosphor layer. _-x Sr _x S (calcium sulfide), YAG-SiN (YAG-tantalum nitride) and related crystalline matrix matrices. Various polyoxo pastes can be used to form a binder based on a ruthenium-based phosphor layer. The additional material used in the encapsulated HBLEDs comprises an AlGaInP active film on a GaAs or InP substrate for use by a manufacturer of red LEDs instead of phosphors to produce a red portion of the white light spectrum. In flip chip applications, ceramic based submount substrates are also generally used. Quartz glass is also commonly used to provide a hermetic seal to the package. Lumileds has been introduced wherein the high temperature of the sintered ceramic eliminate Lumiramic ^TM phosphor layer poly silicon oxide as the need of a binder.

In one embodiment, the above materials are assembled onto a submount at the wafer level and tested before they are singulated, cut, and trimmed for grading. The wafer level test can include measurements of forward voltage resistance, luminous efficacy, CCT, and color spectrum (ie, CIE coordinates) under operating conditions of the die under conditions in which the final LED product will be used. Ideally, the color spectrum will be measured within a MacAdam ellipse.

One of the difficulties in thermal testing mentioned above, which is related to the first one, is the production of product grade HBLEDs at wafer level testing (ie, prior to incorporation into the final lighting product). Difficulties in operating conditions. As explained above, in a product grade packaged HBLED phosphor, its various materials are operated at substantially different temperatures. For example, driving an exemplary InGaN film with a current of 900 mA can achieve 85 ° C or more above room temperature. This is the result of the efficiency of the InGaN film itself on the product grade HBLED and one of the thermal resistivity from the InGaN film through the submount and to the heat sink (eg, the extruded aluminum fins that are convectively cooled). On the other hand, on the surface facing away from the InGaN film, a phosphor layer having a polyfluorene as a binder is operated at a much higher temperature (for example, about 200 ° C). In production grade HBLEDs, the heat sink allows operation at these high temperature conditions. FIG. 6 illustrates an exemplary product grade HBLED including one of the InGaN films 604 encapsulated by a lens 601 prior to attachment to a heat sink. A thermal pad 605 is attached underneath a submount 606 and in electrical contact with the InGaN film 604 for attachment to a region of a heat sink. An anode 602 and a cathode 603 are also shown. Both the anode 602 and the cathode 603 are electrically connected to the InGaN film 604. And form part of the wafer level HBLED.

The high temperature experienced by the top surface of the phosphor layer is approximately one hundred to five hundred less than that of the other materials in the HBLED due to the formation of the doped oxygen-based phosphor layer and the lens of the lens. One of the thermal conductivity. Similarly, when other materials reach equilibrium operating conditions in approximately 10 milliseconds to 20 milliseconds, the thermal time constant of the polyoxyl lens results in a slow 24/25 (the thermal diffusion length is proportional to the square root of the material diffusion constant) and in fact It takes about one second to two seconds. These significantly different operating temperatures and equilibration times present significant challenges to the thermal testing of HBLEDs.

To measure a wafer level packaged HBLED at the correct phosphor equilibrium temperature profile (i.e., prior to attachment to a heat sink), each InGaN film and adjacent phosphors need to be heated for 1 second to 2 seconds. Unfortunately, there is no thermal mass sufficient to achieve this heating at this stage of the process because the InGaN blue-light film also reaches the elevated temperature of the phosphor. (Although this in fact limits this method to heating the entire brick to 85 ° C (the final product operating temperature of the InGaN film in the final SSL product), this leaves the phosphor at a much lower level than the final product. At the temperature of the temperature). Therefore, the absence of an extruded aluminum heat sink (or similar heat sink) for each die also poses a significant challenge to the thermal testing of wafer-level packaged HBLEDs.

To perform the thermal test, a current is applied to the InGaN film (for example, using anode 602 and cathode 603 of FIG. 6) during CIE coordinate measurement to provide electroluminescence from the InGaN junction. (Note that one of the photoluminescent system films is known to be an inappropriate substitute because its condition is significantly different from the electroluminescence response). Although applying 900 mA current to the InGaN film for 20 milliseconds will Heating to approximately 85 ° C, but the phosphor layer and lens are not close to their product grade operating conditions. In fact, the only heat reaching the phosphor layer and the lens is due to the Stokes shift from the (slow) heat conduction of the InGaN film through the phosphor layer and the phosphor luminescence process. Thus, the phosphor layer only reaches thermal equilibrium after one to two seconds of optical excitation and provides its true color spectrum (ie, the spectrum provided in the product grade HBLED).

Note that applying the electroluminescence excitation to the phosphor for only two seconds for two reasons does not allow for a solution. First, for a four inch wafer with 10,000 LEDs, this excitation will take more than 2 hours to test. LED manufacturers need to be faster than this speed from ten times to even forty times the inspection speed to be cost effective. Second, and more importantly, supplying current to the InGaN film for two seconds while not attaching to a heat sink causes the InGaN film to reach such a high temperature that it self-quenches one of the operating temperatures due to thermal flipping, thereby ensuring Error CIE coordinate measurement.

Preheating a wafer-level packaged HBLED in a furnace at 200 °C to achieve the proper peak temperature of the phosphor in the final product operating conditions is also not a viable solution because no correct thermal gradient within the phosphor layer is produced. . Specifically, the temperature in the phosphor layer is not a uniform temperature, but firstly it is a linear gradient from 200 ° C or more at a surface farthest from the InGaN film to 85 ° C at a surface in contact with the InGaN film. . Gradient for the formation of the phosphor layer with one Lumiramic ^TM For less severe, but still significant. For example, Figure 7 illustrates a self based on poly-silicon oxide at the bottom of the phosphor layer (dashed line) and a phosphor layer on one Lumiramic ^TM one (solid line) product level HBLED area 703 of the ceramic submount, after a phosphor The bulk region 702 (for the context provided InGaN film region 703A) reaches an exemplary temperature gradient at the top of the lens region 701. In the FIG. 7, at the interface region 701 of the lens, the phosphor layer based on silicon up to a temperature exceeding 200 ℃ one, but based on the Lumiramic ^TM phosphor layer may reach more than one temperature 100 deg.] C. This maximum temperature of the phosphor layer region 702 undergoes a substantially linear gradient at the interface of the ceramic region 703 to a reduced temperature of 85 °C.

Note that the deviation from the linearity exhibited in the phosphor layer region 702 may exist due to uneven doping within the phosphor layer. In particular, the phosphor crystal distribution within the phosphor layer is generally non-uniform. The regions near the phosphor crystals (although the regions have a diameter of about a few microns, but different among the manufacturers) will be hotter than the phosphor microcrystals, which are substantially non-uniformly disposed on the phosphor layer. Inside. Thus, the uneven distribution of the phosphor crystals creates a hot spot that typically varies between LEDs. Note that only the actual heat source in the phosphor layer is the active ions in the phosphor crystal. Therefore, any hot spot test should be able to reproduce the hot spot temperature as closely as possible.

Lens area 701 serves to protect the die from moisture and to help extract light from phosphor layer region 702. When the lens area 701 does not contain any light-emitting elements, it acts as an insulator. In contrast, the InGaN film 703A on the ceramic sub-substrate region 703 acts as a heat conductor that is opposite to the behavior of the lens region 701. Each of lens area 701 and phosphor layer area 702 exhibits a linear temperature gradient (but different temperature gradients, as shown in Figure 7).

Note that the wafer-level packaged HBLED is initially preheated to a slightly higher temperature. And then the temperature is maintained for a period of 1 second to 2 seconds by applying electroluminescence from the InGaN film, which results in a thermal test of one of the InGaN films at 85 ° C, but the phosphor remains The phosphor layer did not exhibit the correct thermal gradient conditions at the correct elevated temperature. Therefore, a more flexible approach is required to achieve substantially different temperature and temperature profiles within extremely different materials.

As indicated above, it is important to obtain the correct junction temperature for the HBLED for each wafer level package and the correct phosphor layer temperature and temperature gradient. For each purpose, the thermal test should provide conditions that closely simulate the actual operating temperature and temperature gradient to provide a consistent dense grading bin, which is expected to meet customer lighting applications.

FIG. 8 illustrates one exemplary thermal testing technique 800 that can provide accurate high throughput testing of HBLEDs for wafer level packaging. In step 801, a laser can be used to selectively heat portions of the phosphor layer. Notably, by using selective laser heating, the correct temperature gradient in the phosphor layer can be quickly generated in an unbalanced manner, ie, the correct temperature exists before it is degraded by diffusion into adjacent layers of the HBLED package. The lens or adjacent layer (such as an InGaN film) is also not heated for a period of time. In one embodiment, the laser heating can include the use of medium infrared (IR) radiation (eg, a wavelength of 850 nm to 900 nm) to heat a polyoxynitride-based phosphor layer. Note that the only organic material in the polyoxygen-based package is therefore the only material that absorbs optical radiation in the IR region of the electromagnetic spectrum. All other materials used in the packaged HBLED are completely transparent in this area. By means of intentional introduction as an additional material in the binder The combined vibration mode of various types of polyoxymethylene or other organic materials absorbs optical radiation in the region between 3.2 microns and 3.4 microns. In one embodiment, the wavelength of an IR tunable laser can be tuned to select an absorption depth anywhere from a few microns to over 100 microns in the binder material.

In another embodiment, this laser heating can include the use of visible radiation to directly excite active phosphor ions within the matrix crystal. Note that this embodiment can be used in any type of phosphor layer (e.g., based on the phosphor layer of poly-silicon oxide or based on the phosphor layer Lumiramic ^TM). Lumiramic ^TM based on the case of the phosphor layer, it is necessary to use a separate direct optical excitation source to the phosphor, thereby effectively as one of the underlying InGaN film surrogate. An exemplary laser can include an optically pumped semiconductor (OPS) laser that operates at a selected wavelength between 390 nm and 750 nm. In one embodiment, one of the 450 nm wavelengths can be targeted. Other exemplary lasers include an InGaN laser diode array or a dye laser array.

In step 802, an appropriate current can be applied to the InGaN film, thereby providing the correct temperature quickly at the junction (i.e., the junction of the InGaN film). Specifically, since the thermal response time of the interface of the InGaN film is about 100 microseconds, the film can reach the product-level operating temperature in this time due to the current applied to the film during this time. The waste heat is sufficient to cause the correct temperature to rise to 85 ° C at the junction of the heat capacity of a given membrane. In the case of both the correct temperature gradient and the correct junction temperature in the now established phosphor layer, photometric measurements can be made in step 803 for the wafer level packaged HBLED. Notably, these may be measured (either polyethylene oxide or based on silicon-based Lumiramic ^TM) of the retention period in which the time constant which is a predetermined phosphor material.

Notably, selective heating using a laser advantageously decouples the heating time, heat capacity, and thermal time constant of the phosphor layer from the InGaN film and lens regions. Thus, selective laser heating allows for the formation of fast (high throughput), accurate and well defined temperature gradients within the phosphor layer. These features of laser heating allow a manufacturer or other user to conveniently vary and map the temperature distribution and final color coordinates relative to a range of customer operating conditions. Therefore, it is believed that an ordered HBLED is performed based on the final package operating conditions.

Selective laser heating can also be done very quickly. For example, a simple calculation of the heat capacity of a roughly 65 micron thick polyoxo layer shows that the polydecyl oxide can be heated to a median of the z operating point in less than 10 milliseconds with only 10 mj of optical IR radiation. Or the middle range is 180 degrees Celsius (or any temperature expected by the manufacturer). Assuming a laser with an average power of 12 watts and operating at 1 kHz, up to 1200 dies can be processed per second (ignoring any optical losses in the test tool). Therefore, when the equivalents each have four 10,000 wafers or bricks of 10,000 HBLEDs, more than 400 wafers or bricks can be measured per hour if all measurement operations are completed in 10 milliseconds. In practice, the integration time of the spectrometer may be in the range of 20 milliseconds or more and there is additional time to step from the die to the die.

FIG 9 illustrates a phosphor layer on the Lumiramic ^TM exemplary timing and sequence of the phosphor layer of a polyethylene oxide based on silicon. At a time T1 indicating the start of one of the thermal tests of the HBLED for a wafer level package, the portion of the phosphor layer of the HBLED is selectively heated. The period of time required to perform this selective heating depends on whether the material to be heated is polyoxo (in this case, the desired period of time is about 1 millisecond) or the active ions in the phosphor itself (in this case, The required time period is less than approximately 100 microseconds). At time T2, approximately 250 microseconds is provided to allow a suitable temperature gradient to propagate through the phosphor layer, and then a current is applied to the InGaN film for a period of time ranging from about 50 microseconds to 100 microseconds, which brings the InGaN film to 85 °C. At time T3, the light metric is taken for 1 millisecond or more. At time T4, the next wafer level packaged HBLED is positioned for thermal testing, conservatively estimating that the thermal test takes between 20 milliseconds and 40 milliseconds. Taking into account the above, a thermal test tool can be able to inspect and properly grade nearly 100 four-inch wafers or bricks per hour, which is 10 times faster than any other currently commercially available probe test tool. Advantageously, this reduced inspection and grading test time will significantly reduce the manufacturing cost of the thermal test operation.

Note that the temperature gradient provided by selective laser heating is unbalanced because it has not yet heated the boundary of the surrounding material to a comparable temperature. Therefore, preferably, the CIE coordinates are performed before the thermal gradient distribution is modified by conducting the deposition heat to the adjacent lens (upper) and the InGaN film (below) or before the heat transfer balances the temperature distribution within the phosphor layer itself. Measure. A time-dependent model has been used to calculate polyphosphoric acid binder (diffusion rate of 1.3 × 10 ^-7 m ² /s) and Lumiramic ^TM phosphor (one of 4 × 10 ^-6 m ² /s assumed diffusion rate as an upper limit) The diffusion distance of heat transfer within the phosphor layers of both. For the phosphor μ, heat curves a thickness of about 200 for silicon oxide and poly structure for Lowering Lumiramic ^TM beyond this one size fraction (20%) in 0.1 ms and 2.5 ms in. Therefore, preferably, the CIE coordinates should be made within this limited time after the volume change curve is established. Note that the web is applicable regardless of whether the phosphor layer, polyfluorene oxide, (middle IR) or active phosphor (visible excitation) itself is excited.

In an embodiment of selective laser heating of the phosphor layer of the phosphor layer, a temperature gradient is achieved by exciting the polyfluorene oxygen at the appropriate wavelength, wherein the coupling close to the IR radiation is controlled to achieve an almost identical distribution. Figure 10 illustrates a graph 1000 of one of the linear gradients expected when operating a HBLED lamp with cw (continuous wave) compared to the exponential decay and absorption of medium IR radiation. Notably, the two curves do not deviate from each other by more than 3 °C. The total median or average temperature can be achieved by controlling the number of millijoules in the exposure, and the correct temperature gradient can be achieved by operating the laser at the correct wavelength that matches the proper absorption profile of the polyoxyxene material being used. .

Note that the polyoxygen oxide in the phosphor layer may have a crosslinked linkage based on a vinyl group, a phenyl group, a methyl group (eg, PDMS (polydimethyl methoxy oxane)), and related structures, each of the structures A unique medium IR absorption, refractive index, CTE and related mechanical properties are imparted to the binder. Figure 11 illustrates near IR (12000 cm ^-1 to 4000 cm ^-1 (0.8 μm to 2.5 μm wavelength)) absorption and some intermediate IR (4000) of PDMS with basic and specific combination bands labeled with their respective characteristics. Cm ^-1 to 3000 cm ^-1 (2.5 μm to 3.3 μm wavelength)). The spectrum of PDMS at a thickness of 1 cm is labeled as line 1101, and the spectrum of PDMS at a thickness of 0.2 mm is labeled as line 1102. The highest energy vibration mode is the CH (hydrocarbon) bond present at 3.3 microns. This strong absorption feature will also have a companion band that is close to the maximum of the laser gain curve at 2.5 microns due to the combination of the R-Si bending mode at 11.7 microns and the stronger mode at 13 microns. The extinction coefficient of these and other polyoxygen vibration absorption bands can be extracted. As a result, the IR characteristic of the polyoxygen has an absorption depth close to several tens of micrometers for CH stretching and an absorption depth closer to several micrometers for a longer wavelength basic belt.

The various vibrational bands of different polyoxo can be displaced from each other by energy up to several hundred cm ^-1 . Figures 12A and 12B show near IR spectra and some intermediate IR spectra of two polyoxo types (i.e., methyl-vinyl and phenyl, respectively). Figure 12C shows the near IR spectrum and some of the intermediate IR spectra of an exemplary lens sample made from one of the PDMS blue shifts along with the phenyl-based polyfluorene together with the vinyl-based polyfluorene. This means a shift selective excitation of the phosphor layer based on the Lumiramic ^TM volts below the phosphor required: In the case of the above it is not heated above the dome lens (overdome lens) will need to be longer than the oscillation wavelength to one wavelength substantially The underlying phosphor is selectively excited. It is also noted that the combined strip of Figures 12A and 12B has an absorption intensity that is too small by about one-half for a 100 μ thick structure used to provide the desired temperature gradient shown in Figure 7.

In one embodiment, the phosphor layer can be doped with approximately one percent of methanol, which in the case of a thick phenyl polyfluorene lens can provide the possibility of selective excitation in IR in polyfluorene oxide. Sex. Methanol will not affect the properties of polyfluorene (except for wetting onto the substrate) and has OH (oxygen-hydrogen) stretching at 3682 cm ^-1 in the vapor phase and 3400 cm ^-1 in the liquid phase An important feature of this OH stretching is that the basic CH stretching is 500 cm ^-1 away from all polyfluorene oxygen. The use of a 2% wetting solution when applying a phosphor layer produces a strong and selective excitation of a layer of light having an absorption depth of about 50 microns to 100 microns (or even a thin layer at a higher concentration). ability. Evidence suggests that methanol does not affect the opacity, refractive index, chemical, thermal or mechanical properties of the matrix polyfluorene. In addition, methanol can diffuse from the sample within a few hours without affecting LED performance.

As described above, in one embodiment, the activity of the direct excitation of the phosphor layer, ions can be based on a non-polyethylene oxide of silicon (such as using their layer of sintered ceramic (e.g., based on the phosphor layer Lumiramic ^TM)) and A polyphosphonium-based phosphor layer. Notably, the sintered ceramic is as transparent as the rest of the HBLED in the medium IR. Figures 13A, 13B and 13C show the variation of Lumiramic ^(TM) plate thickness and its effect on Δu'v' and CCT. In particular, Figure 13A shows the u'v' CIE 1976 coordinate that can be obtained with different blue-emitting LEDs (shown by dashed lines) and different Lumiramic ^(TM) plate thicknesses (shown by solid diamonds). FIGS. 13B and FIG. 13C shows the plate thickness Lumiramic ^TM different the LED emits blue light (shown as a solid line) obtained from the sensitivity and the Planckian locus (shown as a dotted line in FIG. 13B) of △ u'v 'bias and A CCT of Ce:YAG phosphor (shown as a dashed line in Figure 13C) was used. These changes and sensitivities support the thermal testing of HBLEDs to accurately grade CIE color coordinates within a MacAdam ellipse.

Direct excitation of the phosphor active plasma heating Lumiramic ^TM phosphor and a phosphor of a crystalline matrix material. An exemplary heating source to accomplish this excitation can include an OPSL (optical pumped semiconductor laser) having a wavelength settable (eg, from 350 nm to 600 nm). In one embodiment, OPSL and associated phosphors can be set using a wavelength of 460 nm centered near the strong absorption features of Eu ^{+ 2} , Ce ^{+ 3} as shown in FIG. Therefore, OPSL can simulate the excitation of blue InGaN LEDs. However, in the preferred embodiment, the OPSL is only used to prepare the temperature gradient of the phosphor active ions and is not used to provide the desired junction temperature of the InGaN grains.

Figures 14A and 14B show the excitation spectra of two exemplary phosphors in the form of several crystalline matrix types. Specifically, FIG. 14A shows an excitation spectrum of (Ca _1-x Sr _x )S:Eu+2 having different Sr:Ca ratios, and FIG. 14B shows (Y _0.97 Ce _0.03 ) ₃ Al ₅ O ₁₂ (λ _cm =560nm; the excitation spectrum labeled YAG) is compared to (Y _0.97 Ce _0.03 ) Al _4.9 Si _0.1 O _11.9 N _0.1 (λ _cm = 720 nm; labeled YAG-SiN) in the coordination layer of Ce ³⁺ The room temperature excitation spectrum of the ions.

Note that within one OPSL (between 460 nm and 530 nm) wavelength stability (as opposed to tunability), the absorption profile of the phosphor changes by more than an order of magnitude. Since the excitation dose provides an average temperature and the excitation coupling or absorption profile determines the temperature gradient, phosphor excitation can advantageously provide significant flexibility to accurately set the phosphor temperature gradient across z, and the phosphor temperature gradient on z is achieved Proper thermal test conditions and CIE coordinates are key.

In summary, a variety of phosphors may be used to heat the heating strategy test HBLED having phosphor layers Lumiramic ^TM of the phosphor layer based on polyethylene oxide of silicon or based. The trade-off between the IR spectrum excitation of a polyoxynene binder and the visible spectrum excitation of a reactive ion phosphor is a tunable mid-IR excitation method that uses a more versatile visible spectrum to directly excite the absorption profile flexibility of the phosphor active ions (and The temperature gradient that can be achieved) provides flexibility to cover all types of phosphor packages.

As explained above, the subsequent use of current and rapid photometric extraction of electroluminescent applications with precision laser non-equilibrium heating provides the HBLED industry with a flexible, accurate and high throughput thermal test method.

Note that one of the additional factors affecting the lateral (i.e., parallel to the plane of the quantum well) light distribution of both blue InGaN emission and phosphor emission can be strongly modified by Mie scattering in the phosphor region. Since the wavelength of InGaN pumping radiation can generally be close to the size of the phosphor microcrystals, this means that the direct phosphor pumping condition can be such that its lateral thermal deposition amount curve is modified by the Mie scattering in the lateral dimension. Mie scattering is also present due to IR excitation in the polyoxynene cement pumping process, but it is less severe.

When the size of a particle is comparable to the wavelength of light that traverses the medium, Mie scattering illuminates the scattering of electromagnetic radiation by a sphere or other particle shape. The well-known approximate value of Rayleigh is the intensity I of the scattered radiation given by:

Wherein I ₀ is the intensity of one of the unpolarized beams of wavelength λ, the refractive index of the n-type particles, the diameter of the d-type particles, and the distance of R from the particles.

Therefore, light scattering should be considered when using wavelengths different from those used to excite phosphors in product-level HBLED phosphor excitation (ie, wavelengths are substantially different from those of InGaN electroluminescence). . This scattering is similar in the case of a thermal test by direct visible excitation of the phosphor, since the thermal wavelengths of the provided thermal test source and the electroluminescent source are very similar. However, attention should be paid to the polyoxygen-based phosphor thermal test in which one of the IRs is used instead of a blue wavelength of light.

Figure 15 illustrates the change in Mie scattering coefficient versus wavelength for particles of various sizes. Mie scattering for thermal testing can provide a lateral temperature gradient that is different from one of the temperature gradients that would be obtained from cw electroluminescent illumination conditions. Figure 15 indicates that the Mew scattering coefficient is close to the absorption coefficient in the phosphor only when the average phosphor particle size is one micron or more and for an absorption length of up to one micron of the particle of up to one micron. Therefore, the overall effect of Mie scattering on CIE measurements should be manageable.

Figure 16 illustrates an exemplary thermal test system 1600 for testing of HBLEDs for wafer level packaging. An excitation laser 1602 is provided to excite portions of the phosphor or phosphor layer and establish an appropriate temperature gradient therein. A probe tester 1606 provides current to the HBLED 1607 to bring the InGaN film to 85 °C. In one embodiment, the excitation laser 1602 and probe tester 1606 are controlled by timing electronics 1601 to provide for a suitable period of time for laser excitation and current application.

An integrating sphere 1604 (also known in the industry as Ulbricht sphere) having one of the inner surfaces of the light uniformly scattered at all angles facilitates the collection of light from the HBLED 1607 after laser excitation and current application. The integrating sphere 1604 consists essentially of an optical element consisting of a hollow spherical cavity having one of the apertures for the inlet and outlet. In one embodiment of the integrating sphere 1604, the inlet can include a ring 1604A that is angled to provide a tight fit around the lens of the HBLED 1607 during thermal testing, thereby ensuring that light exiting the HBLED 1607 is not collected. Ring 1604A may include an integrating sphere 1604 A high angle reflective optic that collects light from HBLED 1607 at an angle from 10° to 170°. In one embodiment (shown in FIG. 16), the beam from the excitation laser 1602 can be directed through the integrating sphere 1604 to the HBLED 1607. In other embodiments, the beam of light may be directed obliquely onto the HBLED 1607 without passing through the integrating sphere 1604.

One of the sensors 1603 located at the exit of the integrating sphere 1604 can collect substantially all of the light incident on the inlet and provide the sum of the incident light to a spectrometer system 1605. The spectrometer system 1605 can include a spectrometer and other well-known components for performing light metrology measurements of light from the HBLED 1607, such as a computer. In one embodiment, timing electronics 1601 can control spectrometer system 1605, thereby allowing synchronization with the timing of excitation laser 1602 and probe tester 1606. In one embodiment, the sensor 1603 is turned off when a phosphor non-equilibrium temperature profile is generated during application of the laser, and is only turned on during application of the electroluminescent excitation radiation from the HBLED 1607.

In one embodiment, a wafer carrier 1608 can hold the HBLED 1607 and other HBLEDs in place during thermal testing. In some embodiments, wafer carrier 1608 can be coupled to a conventional mobile platform system 1609, thereby allowing thermal testing of each HBLED on wafer carrier 1608. In one embodiment, timing electronics 1601 can also control platform system 1609 in the x, y, and z planes.

Although the illustrative embodiments of the present invention have been described in detail herein with reference to the drawings, it is understood that the invention is not limited to the precise embodiments. It is not intended to be exhaustive or to limit the invention to the precise form disclosed. As such, many modifications and variations will be apparent to those skilled in the art. For example In other words, although HBLEDs are set forth herein, phosphor-converted HBLEDs (pc-HBLEDs) can be tested in substantially the same manner using substantially the same system. Moreover, although optically pumped semiconductors are set forth herein, other lasers (such as dye lasers or InGaN lasers) may be used in the applications set forth above for optical pumping semiconductors. Accordingly, the scope of the invention is intended to be defined by the scope of the claims

100‧‧‧High-brightness light-emitting diode

101‧‧‧Thin Indium Gallium Nitride Film/InGaN Film/Thin InGaN Film

102‧‧‧phosphor layer

103‧‧‧ lens

104‧‧‧Sub Substrate

601‧‧ lens

602‧‧‧Anode

603‧‧‧ cathode

604‧‧‧InGaN film

605‧‧‧heat pad

606‧‧‧Sub Substrate

701‧‧‧ lens area

702‧‧‧ Phosphor layer area

703‧‧‧Mask sub-substrate area/ceramic area

703A‧‧InGaN film area/InGaN film area

1600‧‧‧ Thermal Test System

1601‧‧‧Timed electronics

1602‧‧‧Inspired laser

1603‧‧‧Sensor

1604‧‧·score ball

1604A‧‧‧ Ring

1605‧‧‧ Spectrometer system

1606‧‧‧Probe Tester

1607‧‧‧High-brightness light-emitting diode

1608‧‧‧ Wafer carrier

1609‧‧‧Useable mobile platform system/platform system

T1‧‧‧ time

T2‧‧‧ time

T3‧‧‧Time

T4‧‧‧ time

Figure 1 illustrates an exemplary wafer level package of HBLEDs.

2A and 2B show plots of decay time vs. temperature for phosphors using two different transmit filters.

Figure 3 illustrates the change in CIE coordinates as a function of the phosphor doping level of a sample having different weight ratios of green phosphor to red phosphor.

Figure 4 shows the CIE variation of one exemplary phosphor binder in the case where a phosphor is applied by a mechanical dispenser.

Figure 5 shows a plot of CIE x versus CIE y plotting a phosphor mixture that is subjected to various current and phosphor plate temperatures.

Figure 6 illustrates an exemplary product grade HBLED.

Figure 7 illustrates an exemplary temperature gradient in an exemplary wafer level packaged HBLED.

Figure 8 illustrates an exemplary thermal testing technique for a HBLED for a wafer level package.

Figure 9 illustrates an exemplary timing sequence for a phosphor. Selective heating is applied for one cycle from 1 microsecond to 1 millisecond to avoid laser damage to the material.

Figure 10 illustrates a graph showing one of the phosphor region heat curves obtained by trimming the absorption depth by tuning the laser to the correct absorption profile.

Figure 11 illustrates the near IR absorption of PDMS and some intermediate IR.

Figures 12A, 12B, and 12C show near IR spectra and some intermediate IR spectra of three different types of polyfluorene oxide.

Figure 13A shows an exemplary U'V' CIE 1976 coordinate that can be obtained with different blue-emitting LEDs and different Lumiramic ^(TM) plate thicknesses.

Figures 13B and 13C show the sensitivity of Lumiramic ^(TM) plate thickness to different blue-emitting LEDs and the resulting deviation from Planck's trajectory Δu'v' and CCT using Ce:YAG phosphors.

14A and 14B show excitation spectra of several exemplary phosphors of various crystalline matrix types.

Figure 15 illustrates the change in Mie scattering coefficient versus wavelength for particles of various sizes.

Figure 16 illustrates an exemplary thermal test system for testing of HBLEDs for wafer level packaging.

Claims (27)

A method of performing a thermal test of a high-brightness light-emitting diode (HBLED) comprising an indium gallium nitride (InGaN) film and a phosphor layer formed on the InGaN film, the method comprising: using a Selectively heating a portion of the phosphor layer to provide a predetermined temperature gradient in the phosphor layer; applying a current to the InGaN film to establish a predetermined temperature in the InGaN film; and performing a photometric measurement on the HBLED. The method of claim 1, wherein the selectively heating directly heats the polyoxane in the phosphor layer. The method of claim 1, wherein the selective heating is performed with a mid-infrared (middle IR) laser. The method of claim 1, wherein the selective heating is performed with a co-ordinated laser. The method of claim 1, wherein the selective heating directly heats the active phosphor ions. The method of claim 1, wherein the selective heating is performed with an InGaN laser to excite an absorption band of approximately 460 nm. A system for thermal testing of high brightness light emitting diodes (HBLEDs), each HBLED comprising an indium gallium nitride (InGaN) film and a phosphor layer formed on the InGaN film, the system comprising: a laser Relocating to direct its light onto an HBLED configured to selectively heat portions of the phosphor layer; a probe tester configured to apply a current to the InGaN film of the HBLED to establish a predetermined temperature and provide electroluminescence in the InGaN film; an integrating sphere configured to collect during testing Light emitted by the HBLED; and a spectrometer system configured to perform photometric measurements on light collected by the integrating sphere. The system of claim 7, further comprising timing electronics coupled to the laser and the probe tester to synchronize the laser and operation of the probe tester. The system of claim 7, wherein the laser is positioned to direct its light through the integrating sphere to the HBLED. The system of claim 7, wherein the integrating sphere comprises a loop configured to minimize extraneous light to the entrance in the integrating sphere during the test and to collect all of the light emitted by the HBLED. The system of claim 7, further comprising a movable carrier for locating the HBLED. A method of performing a thermal test of a high-brightness light-emitting diode (HBLED) comprising an indium gallium nitride (InGaN) film and a phosphor layer formed on the InGaN film, the method comprising: using a first An excitation source to establish one of the first predetermined operating conditions of the phosphor layer; using a second excitation source to establish a second predetermined operating condition of the one of the InGaN films; A photometric measurement is performed on the HBLED after establishing the first predetermined operating condition and the second predetermined operating condition. The method of claim 12, wherein establishing the first predetermined operating condition comprises providing a predetermined temperature gradient for the phosphor layer, and establishing the second predetermined operating condition comprises providing a predetermined temperature for the InGaN film. The method of claim 12, wherein the using the first excitation source comprises: stimulating excitation of polyfluorene oxide used as one of the binders in the phosphor layer. The method of claim 12, wherein the using the first excitation source comprises: exciting the excitation of the active phosphor ions in the phosphor layer. The method of claim 12, wherein the using the first excitation source comprises: using an optical source to selectively excite the vibration mode of the polyoxygen in the phosphor layer, thereby creating a temperature gradient in the phosphor layer. The method of claim 12, wherein the using the first excitation source comprises: using an optical source to selectively excite a vibration mode of one of methanol and a hydrocarbon wetting agent in the phosphor layer, whereby the phosphorescence A temperature gradient is created in the body layer. The method of claim 12, wherein the second excitation source comprises applying a current to the InGaN film. A system for thermal testing of high-brightness light-emitting diodes (HBLEDs), each HBLED comprising an indium gallium nitride (InGaN) film and a phosphor layer formed on the InGaN film, the system comprising: a first An excitation source configured to establish a first predetermined operating condition of the phosphor layer; a second excitation source configured to establish one of the InGaN films a predetermined operating condition; an integrating sphere for positioning over the HBLED, the integrating sphere configured to collect light emitted by the HBLED during testing; and a spectrometer system configured to pair the integral The light collected by the ball performs a light measurement. The system of claim 19, wherein the first excitation source comprises one of an optical parametric oscillator and a Cr+3 insulated crystal laser, the first excitation source configured to be used as the phosphor layer A binder of polyoxyl. The system of claim 19, wherein the first excitation source comprises an InGaN laser configured to excite active phosphor ions in the phosphor layer. The system of claim 19, wherein the first excitation source comprises a coherent light source configured to excite active phosphor ions in the phosphor layer. The system of claim 19, wherein one of the first excitation sources has a wavelength between 2.0 microns and 3.5 microns, and one of the homologous sources has an average power between 100 watts and 12 watts for selectively exciting the phosphorescence Any one of the bulk polyoxymethylene and methanol doping wetting agents. The system of claim 19, wherein one of the first excitation sources has a wavelength between 0.45 micrometers and 0.53 micrometers, and one of the homologous sources has an average power between 100 watts and 12 watts. The system of claim 19, further comprising a plurality of first excitation sources, the plurality of first excitation sources providing one of 12 watts of average coherent power in combination. The system of claim 19, wherein the second excitation source comprises an electrical probe tester. The system of claim 19, further comprising a plurality of second excitation sources.